A computer implemented technique is disclosed, which may include receiving an image containing text from a mobile computing device at a server, the server having one or more processors. This technique may involve obtaining optical character recognition (OCR) text corresponding to the text at the server, and the OCR text is obtained by performing an OCR on the image. This technique may involve identifying non-textual contextual information from the image at the server, and non-textual contextual information may include (i) contextual information different from the text itself, and (ii) do. Such a technique may include, at the server, obtaining a translation that translates the OCR text into the target language based on the non-text context information, and obtaining the translated OCR text. This technique may include outputting the translated OCR text from the server to the mobile computing device.
展开▼